The stem crustacean Oelandocaris oelandica re−visited

MARTIN STEIN, DIETER WALOSZEK, ANDREAS MAAS, JOACHIM T. HAUG, and KLAUS J. MÜLLER

Stein, M., Waloszek, D., Maas, A., Haug, J.T., and Müller, K.J. 2008. The stem crustacean Oelandocaris oelandica re−visited. Acta Palaeontologica Polonica 53 (3): 461–484.

The arthropod Oelandocaris oelandica from the upper Middle Cambrian “Orsten” of Sweden was recently recognized as a member of the early phase of crustacean evolution based on additional morphological detail from new specimens. Here we present a detailed investigation of all available material. It includes the description of a 400 μm long specimen proba− bly representing an early developmental stage. Variation in size correlated with variation of trunk−segment numbers al− lowed recognition of different instars. The largest specimens do not exceed an estimated length of about 1 mm, indicating that our material may consist only of immature specimens. The characteristic, extremely long antennula of O. oelandica branches into three long rods. It may have served as the major structure to sweep in food, aided by the two subsequent ap− pendages. These and the more posterior limbs were also responsible for locomotion. Minute pores on the outer edges of the posterior limbs and on the trunk tergites possibly contained sensilla originally, which may have served as water−cur− rent detectors. The presence of a minute proximal endite only on the third head appendage suggests a rather basal position of this species within Crustacea, because comparable developmental stages of other known stem crustaceans have such an endite on more of their appendages. Reconstruction of O. oelandica and its life attitudes (referred to the largest instar known) benefited from the application of 3D modelling. These helped, e.g., in identifying the combination of the plesiomorphic feeding function of the antennulae and the specialisation of the exopods of the next two appendages as a step toward the development of a sweep−net mode of feeding, one of the key novelties in the evolution of Crustacea. Such a mode of feeding coupled with locomotion of the three anterior appendages is still practiced in the naupliar and metanaupliar phases of many extant eucrustaceans, and even some adults.

Key words: Crustacea, Arthropoda, morphology, life habits, sweep−net feeding, evolution, phylogeny, stem lineage, computer−aided 3D modelling.

Martin Stein [[email protected]], Department of Earth Sciences (Palaeobiology), Uppsala University, Villavägen 16, 75236 Uppsala, Sweden; Dieter Waloszek [dieter.waloszek@uni−ulm.de], Andreas Maas [andreas.maas@uni−ulm.de], and Joachim T. Haug [joachim.haug@uni−ulm.de], Biosystematic Documentation, University of Ulm, Helmholtzstrasse 20, 89081 Ulm, Ger− many; Klaus J. Müller, Institute of Palaeontology, University of Bonn, Nussallee 8, 53115 Bonn, Germany.

Introduction hexapods) and Eucrustacea as the crown group including all taxa with extant derivatives. In 1983 Müller described six Cambrian putative crustacean Some autapomorphies traditionally proposed for Crusta− species on the basis of specimens in an unusual three−dimen− cea, e.g., differentiation of cephalic appendages into two sional type of phosphatic preservation. Subsequent study of pairs of (sensorial) antennae, mandibles and two pairs of additional arthropod taxa in this “Orsten”−type preservation maxillae, could be soundly refuted and demonstrated to be (Maas et al. 2006) led to the reconstruction of the stem lin− valid only for certain eucrustacean in−group taxa; instead, a eage of Crustacea (putative stem taxa: Cambropachycope new set of autapomorphies was proposed (Walossek and clarksoni Walossek and Müller, 1990, Goticaris longispi− Müller 1990; 1998a, b; Walossek 1999; Maas et al. 2003; nosa Walossek and Müller, 1990, Henningsmoenicaris scu− Waloszek 2003a, b). Two of the new characters are notewor− tula [Walossek and Müller, 1990], Martinssonia elongata thy in this context: Müller and Walossek, 1986; subsequently added: Cambro− – exopods of, at least, the first two post−antennular limbs caris baltica Walossek and Szaniawski, 1991) and identifi− multi−annulated and with locomotory setae arising from cation of the abundant bivalved Phosphatocopina as the sis− the inner side of the annuli facing the endopod; ter group of Eucrustacea (crown group Crustacea; Maas et al. – a small setiferous endite, called “proximal endite”, located 2003; Siveter et al. 2003; Maas and Waloszek 2005). This medially below the basipod of the post−antennular ap− helped consolidate our understanding of the ground patterns pendages (Walossek and Müller 1990). of Crustacea sensu lato, Labrophora (Phosphatocopina + Multi−annulated exopods form part of a sweep−net feed− Eucrustacea, possibly also including the myriapods and/or ing and locomotion apparatus and occur in the Cambrian

Acta Palaeontol. Pol. 53 (3): 461–484, 2008 http://app.pan.pl/acta53/app53−461.pdf 462 ACTA PALAEONTOLOGICA POLONICA 53 (3), 2008 phosphatocopines and in extant small−scale eucrustaceans life attitudes. In addition we describe a small specimen, ap− and all free early eucrustacean larvae. The “proximal endite” parently of an early larval stage as the putative earliest stage had been recognized on extant branchiopods almost 90 years known of this species, filling the series of instars. earlier (Calman 1909), but remained largely ignored subse− quently. We regard its development as a significant step in Institutional abbreviation.—UB, University of Bonn, Ger− the change of the locomotory and feeding attitudes along the many. evolutionary lineage of crustaceans (see also, e.g., Waloszek et al. 2007). Further modification of the “proximal endite” on the second and third appendages, the so−called antennae and Material mandibles, into a coxa is an autapomorphy of the Labrophora (Siveter et al. 2003). Besides these features, the Labrophora All material was etched from bituminous limestone nodules can be distinguished from the stem taxa by more characters (“Orsten”), isolated and mounted on SEM stubs by the work− associated with the feeding and locomotory apparatus, such ing team of K.J.M. in Bonn, Germany in the 1980s to the mid as a fleshy labrum behind the original hypostome, fusion of 1990s. The holotype (UB 649, Fig. 1) of Oelandocaris oelan− the postoral sternites of the antennal to maxillulary segments dica is derived from the Olenus gibbosus Zone (lowermost (sternal plate, sternum), paragnaths on the mandibular por− Furongian Series, Peng et al. 2004; Babcock et al. 2005; tion of the sternum and the appearance of fine setulae or Terfelt et al. 2005) of the Alum Shale succession near Deger− denticles on setae, parts of the labrum, the sternum and the hamn, Öland, Sweden. The additional material is from the paragnaths (Maas et al. 2003). These features are also signifi− Agnostus pisiformis Zone at Gum, Västergötland, Sweden, cant in testing proposed atelocerate/tracheate or hexapod re− now representing the uppermost zone of the Middle Cambrian lationships (e.g., Wolff and Scholtz 2006 recognized homo− (Babcock et al. 2005). The additional material comprises six logy between the paragnaths of myriapods, hexapods and specimens, five larger ones (UB W 260–264, size range from crustaceans). Antennular morphology, development of an− more than 660 μm to ca. 1 mm, Table 2) and a considerably tenna and mandible, exopod morphology, and the occurrence smaller specimen (UB W 265), which is ca. 440 μm long. This and particularly the ontogenetic development of the “proxi− specimen shares characteristic features with the larger speci− mal endite” on postantennular limbs, may eventually facili− mens but evidently represents an earlier developmental stage. tate further resolution of the relationships among the crusta− A brief overview of the preservation of the individual speci− cean stem taxa. mens and features they display is given in Table 1. All compa− Re−investigations of Müller’s original set of taxa (Müller rable features among the specimens correspond in structure, 1983) on the basis of additional material allowed systematic topology, and size. Therefore the material is considered con− assignment of Skara anulata Müller, 1983, Bredocaris admi− specific with the holotype despite the stratigraphical and geo− rabilis Müller, 1983 and Rehbachiella kinnekullensis Müller, graphic distance between Öland and Västergötland. In the ab− 1983 to particular eucrustacean taxa (summary in Walossek sence of diagnostic differences, establishment of a new spe− and Müller 1998b). Walossekia quinquespinosa Müller, 1983 cies for the material from Västergötland is not tenable. and Dala peilertae Müller, 1981 (Müller 1981 treated this Due to the poor preservation of the holotype (Table 1), it species already in sufficient detail to qualify as a valid taxo− is difficult to interpret its morphology well, as exemplified nomic description, accordingly this date has to be used in by the misinterpretation of segmentation and a miscount of preference), awaiting detailed restudy, possess features limbs in the original description (Müller 1983, corrected in known only from particular crown crustaceans such as Stein et al. 2005). The new specimens, though better pre− cephalocarids, maxillopods and branchiopods, considered to served (Fig. 2), are considerably distorted and incomplete. form the monophylum Entomostraca (cf. Walossek 1999). However, the new material increases our understanding of Oelandocaris oelandica Müller, 1983 was originally the holotype, and also allows for comparison of many more known from a single, poorly preserved specimen, but addi− specific structures, such as the shield, the overall shape, and tional material identified subsequently has made a re−investi− the large set of ventral details including the eyes, the hypo− gation tenable. An initial report by Stein et al. (2005) ad− stome, the mid−ventral surface, and, last but not least, the ap− dressed some key features of the new material and concluded pendages. This facilitates reconstruction of Oelandocaris that O. oelandica is a derivative of the stem lineage of oelandica at high fidelity and also helps to understand more Crustacea. Among the crustacean stem derivatives, O. oelan− of the life habits of this tiny early crustacean. dica is unique in having only one “proximal endite”, which Measurements, though difficult to apply due to the distor− occurs on the second post−antennular appendage (the mandi− tion of the specimens (Table 2), revealed considerable vari− ble of labrophoran crustaceans) and remains up to the latest ability. Besides the early instar, three size classes can be dis− developmental stage known so far. The aim of our detailed criminated, which correlate with an increase in the number of re−investigation of O. oelandica, using scanning electron mi− trunk tergites/segments (see section “Ontogeny”). The early croscopy, morphometrics, and computer−based reconstruc− developmental stage, represented by UB W 265, has a total tions, is to present an in−depth report of its morphology, as− length of about 440 μm, with the head region back to the fifth pects of ontogeny, morphogenesis of structures, and possible appendages measuring 290 μm. Set 1 of the larger specimens STEIN ET AL.—CAMBRIAN STEM CRUSTACEAN 463

Table 1. Specimens studied. Abbreviations: app2, 3, 5, appendages; atl, antennula.

UB 649 Coarsely phosphatized; almost complete length, head shield complete, also with rostral extension; specimen is laterally compressed (inwardly pressed shield flanks), shield somewhat distorted dorsally; cephalic appendages preserved in most cases only by their in− sertion sockets (partly with infillings); all distal parts of the appendages missing; only app2 has basipod and proximal part of exopod preserved; some proximal limb parts pressed together; hypostome complete, putative eyes preserved, sternal region well recogniz− able with sternal part of third appendiferous head segment gently elevated; trunk with four segments with short tergites, partly dis− torted; anterior portion of telson with anus on ventral surface preserved as long tubular structure; shield and trunk tergites seem to have borne lateral spines which are broken off, leaving holes on the surface UB W 260 Dorso−ventrally flattened, with trunk flexed against the body; head shield almost completely preserved except the rostrum; putative eyes and hypostome well preserved; three tergite−bearing trunk segments preserved, flattened; proximal portions of antennula pre− served; basipod and 3 exopod articles preserved in app2; basipod and proximal articles of exopod preserved in other limbs; proximal podomere of endopod of app3 present UB W 261 Only left half preserved; head shield missing; 4 tergites and part of caudal end recognizable; only proximal portion of eyes pre− served; posterior part of hypostome missing; antennula preserved with proximal parts of outgrowths, app2 with basipod and 4 exopod articles, of app3 only basipod preserved, subsequent appendages present with basipod and first podomere of endopod UB W 262 Head shield fragmentary; putative eyes and hypostome well preserved; 3 trunk tergites preserved; of atl only proximal portions left; app2 with basipods, endopods and exopods, subsequent appendages with basipods, exopods, and first endopodal podomeres UB W 263 Laterally compressed; anterior and right portions of head shield missing; hypostome and sternal area missing; 5 tergites preserved, tailpiece fragmentary; atl and right appendage series missing; preserved portions of appendages on left side; of app2 only basipod preserved, app3 with basipod and proximal endite within ample joint membrane, proximal endopod podomere, app4 with basipod and first endopod podomere, app5 with basipod, two endopodal podomeres, and part of exopod, first trunk limb with basipod and first endopod podomere; posterior limbs fragmentary UB W 264 Laterally compressed; anterior portions of head shield missing; only few ventral structures preserved, particularly the posterior end of hypostome; also posterior end only fragmentary UB W 265 Head shield and trunk fully preserved; hypostome and left appendage series almost entirely preserved, right appendage series frag− mentary

Table 2. Morphometric data of selected features of the investigated specimens of Oelandocaris oelandica Müller, 1983 (values in μm, values in brackets are estimates; sorted according to the increase in the number of trunk tergites). Total length measured from eyes to start of tail. Head shield measured from eyes. + present but not measured; – absent; ? unclear. Abbreviations: bas, basipod; en, endopod; ts, tergite.

Number of UB W 265 UB W 262 UB 649 UB W 264 UB W 263 UB W 260 UB W 261 tergites (1) 3 4 ³4 5 >3 5 total length 440 >670 740 >685 >795 >945 990 head shield length 290 >430 450 >435 – 590 605 head shield height 90 – 125 ? ? + – eye length 25 – 30 – – 50 – hypostome 120 180 (160) (150) ? ? 220 (195) (210) ts1 – + ? 95 100 110 + ts2 – + 80 85 95 100 + ts3 – + + 60 80 95 100 ts4 – – + + 65 – 80 ts5 – – – ? 45 – 80 tailpiece ? ? ? – ? – ? bas2 height + 60 50 – 65 65 70 bas3 height + + ? – 85 + 100 bas4 height + + ? – 115 110 120 bas5 height – + ? – 120 130 120 bas6 height – 95 ? – 120 100 + bas7 height – – ? – ? 90 + en2.1 height 25 30 ? –––– en3.1 height 30 ? ? –––– en4.1 height ? 55 ? – 55 60 65 en5.1 height ? 60 ? – 60 – 70 en6.1 height – 60 ? – 55 – + en7.1 height – – ? – – – +

http://app.pan.pl/acta53/app53−461.pdf 464 ACTA PALAEONTOLOGICA POLONICA 53 (3), 2008 measures more than 660 μm in total length, with a head the limbs of the model helped to improve the understanding of shield length greater than 430 μm. Set 2 has a total length of the possible range of movement and interoperability in the 680 to 740 μm, with a head shield length between 430 and whole locomotory and feeding apparatus. Later, a refined 3D 450 μm. Set 3 measures up to 1mm in total length, and its reconstruction was created using the modelling software Blen− head shield measures about 600 μm. This size difference of der (MS, JTH). A sequence of a possible limb movement cycle ca. 20–25% between sets 2 and 3 may suggest that more in− was created and the movie is stored in the data repository of stars might have been present, but the segmental increase of Acta Palaeontologica Polonica (see Supplementary Online just one trunk segment does not support this view. The Material at http://app.pan.pl/SOM/app53−Stein_etal_SOM.pdf). holotype falls into set 2. The only major morphological vari− Terminology.—Terminology follows the standardisation of ation between the holotype from Öland and the new material morphological terms for crustaceans and arthropods proposed from Västergötland can be found in the size of the caudo− by Walossek/Waloszek in various papers (e.g. Walossek lateral spines of head shield and trunk tergites. The distance 1993; Waloszek 2003b), which is considered to provide maxi− between the two localities is 330 km, so variation might be mum consistency and terminological stability. Standardisation explained by geographical separation, and possibly also by a is particularly important with regard to segmentation, append− slight difference in age, but the limited material does not al− age morphology and other ventral structures (e.g., differentia− low farther−reaching speculations. tion between “hypostome” and “labrum”, see, e.g., Waloszek 2003b). Since we regard homology of the most anterior seg− mented and pivot−jointed appendages among euarthropods as Methods well established (e.g., Chen et al. 2004; Scholtz and Edge− combe 2005, 2006; Waloszek et al. 2005; but see Budd 2002 Techniques.—Initial SEM investigation of all specimens oc− for an opposing view) we use the term antennula rather than curred in the early 1980s to mid 1990s in Bonn (DW). De− antenna for this appendage, to avoid confusion with the first tailed re−investigations after 2000 were done in Ulm (MS, postantennular appendage in Crustacea, called antenna. The DW) using a Zeiss DSM 962 scanning electron microscope post−antennular limbs are denoted in neutral terminology, i.e., at the Central Facility for Electron Microscopy, University of the “first post−antennular limb” also represents the “second Ulm. All specimens had been glued to SEM stubs prior to our appendage”, and so on. The term podomere is restricted to the re−investigations and could not be re−mounted due to their subdivision of the endopod, but the homology of the subdivi− fragility. Documentation of the morphology in all aspects sions of the antennula and the exopods of the post−antennular was greatly facilitated by an electro−mechanic SEM device limbs are uncertain. Therefore we apply the neutral term arti− constructed by the team of the Central Facility. This appli− cle to these. The terms “anterior and posterior wings” are ance enabled rotation and tilting of the scanned object up to a adopted from trilobite descriptive terminology of the hypo− full angle of 90°, thus extending by far the limited inclination stome, as corresponding structures are present in Oelando− provided by the SEM. caris oelandica and other euarthropods such as e.g., Agnostus Measurements were obtained from the SEM images. Since pisiformis (Müller and Walossek 1987), and the phosphato− high accuracy was not feasible because of preservational limi− copines (Maas et al. 2003). Other terms are explained, as nec− tations (tilting and wrinkling), we adjusted the values to the essary, in the text. Terminology of the armature of the append− nearest 5 μm. This suffices for gross comparisons within the ages was quite difficult to apply due to the fragmentary status. material (Table 2; see also, e.g., Müller and Walossek 1985 Hence we could not clearly distinguish between setae and and Walossek 1993). spines, setulae or spinulae, denticles, etc. Also the size could Reconstruction of some structures remains uncertain due often only be estimated from the insertion area, although the to the only partial preservation of, e.g., limb parts and the tail available “Orsten” forms provide good proxy in estimating end. In some cases, details from the early developmental stage size and form of such cuticular outgrowths. (UB W 265) were used to complete the reconstruction. Where similarity could be estimated, features were also taken from other stem crustaceans, particularly Henningsmoenicaris scu− tula (Walossek and Müller, 1990) (JTH, AM, and DW unpub− Morphology lished data). of Oelandocaris oelandica The SEM images were processed with Adobe Photo− shop™ CS1, the line drawings were produced with Adobe Il− Assignment of the different specimens from Västergötland to lustrator™ CS1 and 3. A plasticine model at 500 times magni− Oelandocaris oelandica rests on the following characters of fication was created in a similar fashion as that of Bredocaris the holotype from the Isle of Öland: admirabilis (Müller and Walossek 1988). Coloured plasticine was used to distinguish different limb parts, adopting the stan− Fig. 1. Holotype of stem crustacean Oelandocaris oelandica Müller, 1983 ® dardized colour scheme introduced by Walossek (1993). We (UB 649) from Degerhamn, Öland, Sweden. A. Lateral view. B. Ventral controlled the proportions using enlarged SEM pictures and view. C. Dorsal view. D. Stereo image of ventral view. E. Latero−ventral line drawings of the appendages. Changing the orientation of view of the head. Abbreviations: app, appendages; ts, tergites. STEIN ET AL.—CAMBRIAN STEM CRUSTACEAN 465

ts1 median eyes head shield ts2 hypostome

antennula

hypostome

head shield

ts1 app2

app3

app4

app5

app6

app7

sternal area hypostome app8

http://app.pan.pl/acta53/app53−461.pdf 466 ACTA PALAEONTOLOGICA POLONICA 53 (3), 2008

– simple elongate head shield drawn out anteriorly into short flat, with the shield margins extending only slightly ventro− rostrum with flat ventral side; laterally. The anterior of the trunk is almost as wide as the pos− – head shield bluntly terminated caudally, most likely with terior end of the head shield and tapers gently posteriorly, ap− lateral spines pointing caudally; parently ending in a longer conical element. This conical ele− – head incorporates five appendiferous segments; ment is only partly preserved, but appears to be similar to the – elongate, slightly elevated hypostome; tail end of Henningsmoenicaris scutula with a medially deep− – paired bulbous protrusions (interpreted as eyes) at front of ened softer area, accommodating the pre−terminal anus and a hypostome; pair of lateral, caudally pointing setae. This is also observed in – nearly circular insertion site of antennula; the early developmental stage (UB W 265) assigned to O. – proximal limb parts of second appendages in holotype. oelandica. The trunk of the holotype comprises four free segments with laterally downward bending tergites extending into short, Head and head shield.—The head shield (Fig. 1) incorpo− ill−defined tergopleurae with pointed or spine−bearing postero− rates five appendiferous segments, the antennular segment lateral corners. The maximum number of segments in the plus four segments with biramous limbs (Figs. 1, 2). Towards larger specimens from Västergötland is five. The general de− the anterior it narrows rapidly, extending into a short conical scription is based on the largest specimens in the material, with rostrum with a bluntly rounded tip (Fig. 1B). The head shield the younger specimens being used as a complement for miss− is moderately arched in cross section and has slightly ex− ing features. tended lateral flanges. These curve down from the rostrum at about the antennulae (Fig. 1A, B) being almost straight later− Oelandocaris Müller, 1983 ally and covering little more of the appendages than the Type species: Oelandocaris oelandica Müller, 1983. body−joint area. The head shield reaches its maximum width between the first and second post−antennular limbs (Fig. 1B). Oelandocaris oelandica Müller, 1983 The flat ventral side of the head is weakly sclerotised (called *v1983 Oelandocaris oelandica sp. nov.; Müller1983: 93, 107; fig. inner lamella in euarthropods and crustaceans with large 11A, B [UB 649], 12. shields, e.g., ostracodes). The only elevated structure on the 1985 Oelandocaris degerhamnensis Müller, 1983; Müller and Walossek ventral side of the head is the elongate hypostome in the mid 1985a: 163 [sic!]. line of the anterior half and the sternal region of the third 2003 Oelandocaris oelandica; Maas et al. 2003: table 2 [holotype spec− imen]. appendiferous segment. The posterolateral corners of the v2005 Oelandocaris oelandica Müller, 1983; Stein et al. 2005: 55–57, head shield curve upward into a straight posterior margin, 60, 62, 64, 67–69; figs. 1 [UB 649], 2, 3A [UB 649], B [UB W which overhangs the anterior trunk tergite only slightly (Fig. 260], C [UB W 261], D [UB W 262], 4A [UB W 261], B–D [UB 2B, E). This gives the head shield a sharply terminating ap− W 263], E [UB W 262], F [UB W 263], G [UB W 260], 5C [UB W pearance in lateral view. Caudolaterally directed spines seem 263], D, 7. to have emerged from the corners. These differ in size indi− 2006 Oelandocaris oelandica Müller, 1983; Maas et al. 2006: 275. vidually, from being very conspicuous (Fig. 2B) to being v2007 Oelandocaris oelandica Müller, 1983; Chen et al. 2007: 264; fig. 11E. only little protrusions (Fig. 2E1). 2007 Oelandocaris oelandica Müller, 1983; Siveter et al. 2007: 2105. v2007 Oelandocaris oelandica Müller, 1983; Waloszek et al. 2007: Hypostome, putative eyes, and sternal region.—The elon− 284; figs. 2B [UB W 261, erroneously labelled UB W 263], C [UB gate, more or less rectangular hypostome extends medially W 263], 3, 5C. from slightly behind the rostrum backwards and ends bluntly at two thirds of the length of the head shield (Figs. 1B, D, E, General habitus 2C, 3). It is straight sagittally, only slightly elevated, and gently sloping laterally. The first post−antennular limbs are The body of the largest developmental stage known of Oelan− inserted at the posterolateral corners of the hypostome (Fig. docaris oelandica comprises two tagmata: head and trunk. 1B, D, E). The anterior hypostomal wings extend to the Both are almost equal in length, the head shield measuring antero−lateral margin of the head shield (Fig. 8B2, arrow). about 600 μm, the trunk about 400 μm (see Table 1; Figs. 1–3) They are overlain by a paired structure, which is composed including a tailpiece estimated to somewhat more than 100 μm of a medially connected peduncle and an ovoid swelling on in length (never complete). The nearly semi−circular cross sec− top of it (Figs. 1B, D, E, 4A, C), possibly representing the tion of both parts and the general appearance —head shield median eyes. Behind the putative eyes, the hypostome is with frontal rostrum−like extension and the segmented tail, slightly swollen. Posterior to that, the hypostomal flanks are most likely extending into a conical end piece—give the ani− slightly constricted to widen again at the posterior margin. mal a shrimp−like appearance. The ventral side of the head is The posterior wings of the hypostome are little pronounced

Fig. 2. Specimens of later developmental stages of the stem crustacean Oelandocaris oelandica Müller, 1983 from Gum, Västergötland, Sweden. A.UBW ® 260, ventral view. B. UB W 264, lateral view. C. UB W 261, lateral view. D. UB W 262, ventral view, specimen slightly distorted. E. UB W 263, lateral view from the left side, left set of appendages broken off, exposing the right limbs (E1) and lateral view of the right side (E2). Arrows point to postero−lateral corners/spines on head shield and tergites. Abbreviations: app, appendages; ts, tergites. STEIN ET AL.—CAMBRIAN STEM CRUSTACEAN 467

head shield app2 antennula

ts1

ts2

ts3

app

app3 exopod

antennula

app4

app2 app3 app2

antennula

hypostome

app4–6

app3 ts1–2

app2 app2

app3

app4

http://app.pan.pl/acta53/app53−461.pdf 468 ACTA PALAEONTOLOGICA POLONICA 53 (3), 2008

rostrum

median eyes hypostome antennula

sternal area

app2

app3

anus telson

Fig. 3. Reconstruction of the morphology of the latest known developmental stage of stem crustacean Oelandocaris oelandica Müller, 1983 in ventral view. Abbreviation: app, appendages. and seem to merge into the antero−median edge of the inser− preserved in the earliest developmental stage (Fig. 9F). From tions of the second limb (Fig. 1B). The mouth opening is lo− the elevation of the hypostome it is clear that the mouth cated below the posterior margin of the hypostome (Fig. 4C). opens above the ventral surface. It seems to be overhung to some degree by the rounded end A sternite belonging to the first post−antennular segment of the hypostome. A mouth membrane, a soft lip−like struc− (antennal segment in labrophoran terminology) could not be ture around the actual opening, may be present, but is only identified. The sternal region between the second post−anten−

Fig. 4. Details of specimens of later developmental stages of the stem crustacean Oelandocaris oelandica Müller, 1983 from Gum, Västergötland, Sweden. ® A. UB W 260; A1, ventral view of anterior head region; A2, latero−ventral view of putative median eyes; A3, close−up of head with antennula and the anteri− orly bent third appendage. B. UB W 261, latero−ventral view displaying antennula. Setal sockets marked by arrows. C. UB W 264, ventral view of head re− gion. Specimen considerably squeezed. Abbreviations: atl, antennula; am, arthrodial membrane; app, appendages; hs, head shield; hyp, hypostome; me, median eyes; ped, peduncle. STEIN ET AL.—CAMBRIAN STEM CRUSTACEAN 469

me me antennula

am

hyp atl hyp

ped atl

app2

hs

me me

app3

hyp 10 µm

atl

app3 hypostome

mouth

sternal area

http://app.pan.pl/acta53/app53−461.pdf 470 ACTA PALAEONTOLOGICA POLONICA 53 (3), 2008 nular appendages may be slightly more sclerotised than the stage (below) and a similar morphology of this region in surface surrounding the appendages, being moderately ele− Henningsmoenicaris scutula (Walossek and Müller 1990; vated (best preserved in the holotype, see Fig. 1E). Towards JTH, AM, and DW unpublished data). The anterior ventral the hypostome the surface is slightly deepened and may be side is slightly depressed and may be pliable, with the anus softer than in its surrounding. A median segmental boundary located in this soft area (Fig. 8C). toward the narrower sternite of the third post−antennular seg− ment is present, but weakly developed, so this sternite is sepa− Appendages rate (Figs. 4B, 5E). The last cephalic sternite is also separate, again narrower than the preceding sternite. The trunk seg− Antennula.—The antennula (Figs. 2C, 7A) inserts directly ments have even narrower spaces between their limbs. There− postero−laterally to the anterior wing of the hypostome (Fig. fore, the food chamber may be limited to the head (Fig. 3). 1B). The insertion area is circular to sub−triangular. The antennula rests on a prominent, socket−like truncated−conical Trunk.—The maximum number of trunk segments encoun− arthrodial membrane (Fig. 4A ). The appendage itself is tered in the largest specimens is five (Fig. 2E, Table 1). Apart 1 composed of a relatively short three−divided peduncle, each from the transversely oriented, elliptical limb insertions, the of the articles carrying a multi−segmented outgrowth disto− ventral surface of the trunk is not sufficiently well preserved laterally (Figs. 4A ,A, 7A). The structure of the peduncle is to allow a detailed description. It seems that the median path 1 3 complex, with many structures involved. All articles are nar− between the limbs becomes progressively narrower towards rowest at their proximal end and widen distally. The first arti− the posterior (Fig. 2A, D). If sternites were present, they must cle is two−divided in the long axis of the appendage, the have been small. smaller lateral portion carrying the outgrowth and a seta on The trunk tergites are of the same shape, except for a de− its anterior surface. The larger median portion is slightly con− crease in size towards the posterior (Fig. 2B, E2). They are stricted proximally and indented postero−distally allowing a roughly U−shaped in cross section. The postero−lateral areas wide flexure of the entire distal part of the antennula and its are slightly imbricated (anterior one overlapping the next). own outgrowth (Fig. 4A3). The subsequent article inserts There are no distinctive tergopleurae (Fig. 2C, E2). The mediodistally on the first article. It is uniform, inverted coni− curved lateral margins are thickened, possibly bearing setae cal as the first article, but smaller in extension and width, car− or sensilla, as is indicated by the presence of corresponding rying a seta on its antero−distal surface, an outgrowth latero− holes and setal sockets in several of our specimens (Fig. 6C1, distally, and the third article mediodistally (Fig. 4B). The see also the chapter on special details below). One prominent third article is barrel to cone shaped, continuing into an out− seta may insert close to the postero−lateral tergal corner, as is growth distally. Even this article bears a seta on its antero− indicated by a socket being larger than the others (Fig. 8B1, distal surface (Fig. 4B). arrow). In a similar fashion as the shield, each tergite is ex− All outgrowths are composed of tubular articles. The first tended into a spine at its postero−lateral corner (Fig. 2B, C). outgrowth consists of at least five articles; the second and An elongate tailpiece is present, but is known only from third outgrowths consist of at least two articles. Presumably fragments in our material of the older developmental stages all three outgrowths are at least half as long as the entire ani− (Figs. 6B2,8B1, D). Part of it is preserved in the holotype, mal, as suggested by additional isolated parts found between where it was reconstructed as a rod in the original description the basipods of the fifth cephalic appendage pair in one of the of the species, though not fully understood (Müller 1983: fig. specimens (Fig. 3). The most proximal article of each of the 12). New photographs show at least the remains of the anus outgrowths is wide at its base and narrows rapidly distally. between the enrolled sides (Fig. 8C, see also Fig. 3), proving The other articles are ca. five times as long as wide and are that this rod−like end is indeed a telson (or a pleotelson de− slightly laterally compressed and oval in cross section. All pending on the number of further segments included in this carry a seta on their antero−distal surface, as is indicated by portion). The anterior portion of the tailpiece is elliptical in the preserved insertion areas (arrowheads in Fig. 4B). The cross section, being slightly depressed in dorso−ventral as− setae may point distally, as is indicated by the position of pect, bearing an indentation on the ventral side (Fig. 8C) and their insertion areas (Fig. 2C). a pair of lateral step−like ridges with attachment sites for setae (Figs. 2C, 8B1 arrow). Presumably the tail is more or Second cephalic appendage (app2).—The first post−anten− less a truncated cone with a flattened ventral side, as is in− nular limb (Fig. 7B) is biramous and is inserted at the postero− ferred from the morphology of the earlier developmental lateral edge of the hypostome. The limb is rotated postero−me−

Fig. 5. Details of specimens of later developmental stages of the stem crustacean Oelandocaris oelandica Müller, 1983 from Gum continued. A. UB W 262; ® A1, lateral view exhibiting the exopod of third appendage; A2, second appendage nestling to the hypostome viewed from the posterior. B. UB W 263; B1,third appendage viewed from the anterior, endopod and tip of exopod missing, small arrows point to setal attachments on median edge of exopod, large arrows mark basipod−exopod articulation; B2, median view of appendages 2 to 4 from postero−lateral, only the arthrodial membranes and the basipods are preserved of the appendages; note the proximal endite of the third appendage. Mirrored for correct view of the mouth area of the animal. C. UB W 261; C1, view from antero−ventral onto the ventral body surface with appendages 2 to 4 and the tips of the antennula; C2, close−up of proximal part of second appendage exhibiting various setae. Arrow points to bifid seta. Abbreviations: alt, antennula; am, arthrodial membrane; app, appendages; bas, basipod; ex, exopod; hyp, hypostome. STEIN ET AL.—CAMBRIAN STEM CRUSTACEAN 471

app4 ex ex

app3 ex hyp app2

app3

proximal endite

spines

bas app2 app4 app4

bas

mouth ex

app3

bas

am

app4 ex

atl

app3 ex

app2 ex

app2

bas

http://app.pan.pl/acta53/app53−461.pdf 472 ACTA PALAEONTOLOGICA POLONICA 53 (3), 2008 dially, with the median edge of the antero−posteriorly flattened The endopod of this limb is unknown except for its articu− basipod being more posterior than the insertion of the exopod. lation site on the medio−distal edge of the basipod (the recon− The arthrodial membrane forms a limb socket. The posterior struction is based on the endopod of the first post−antennular face of the basipod is shorter than the anterior. Along its me− limb, see Fig. 7C). The exopod inserts along the slanting outer dian edge, the basipod carries a bi−serial armature of setae or edge of the basipod and comprises four portions or articles. fine spines that point toward the mouth. The most proximal of The proximal portion is elongatedly paddle−shaped. Proxi− the anterior row is a prominent spine with a bifid tip that mally, it forms a shaft that extends some way down the lateral reaches underneath the distal portion of the hypostome. Distal body wall as part of the body−limb joint (Fig. 5B1). The free in− to it, the anterior series consists of eight equally large setae ar− ner margin carries a series of six setae, opposing the endopod; ranged in a meandering row. The posterior series consists of medio−distally there is a long seta in addition. Posteriorly, the seven setae, the two most proximal of which are smaller than joint between this portion and the basipod is rather wide medi− the distal ones. ally, suggesting the possibility of a wider back swing of the The endopod is attached to the medio−distal portion of the exopod (Fig. 8D). The subsequent two articles are shorter and basipod. It comprises four sub−cylindrical podomeres. Each narrower than the proximal portion and distinctly trapezoidal of the three proximal podomeres widens slightly distally and in anterior view; the elongation of their median faces relative possibly carries a seta on each side of the articulation with the to the exterior faces pre−forms an outward turn of the entire subsequent podomere. The distal podomere is a short conical distal part of the exopod (possibly also reflecting its maximum element probably bearing two setae (Fig. 5A2). The exopod outward flexure, see Fig. 5A1). Medio−distally, the second ar− comprises five antero−posteriorly−compressed articles. With ticle carries one long seta, the third article two slightly smaller roughly double the size of the basipod, it is the dominant setae (Fig. 5A1). The terminal, fourth article is slender, sub− ramus of this appendage. The proximal portion articulates rectangular and carries a set of three setae terminally. along the slanting lateral edge of the basipod (Fig. 1B). The Fourth and fifth cephalic appendages second exopodal portion is shorter, trapezoidal to rectangular .—The third (Fig. in outline. Both portions carried a series of four to five setae 7D) and fourth (Fig. 7E) post−antennular limbs are similar along their median edges, opposing the endopod. Exopodal with one another and their basipods are about 20% longer articles 3 to 5 decrease in size distally. They are trapezoidal than that of the second post−antennular limb (Fig. 2C, E1). in outline and widen distally. All lack setae along their me− The arthrodial membrane forms a prominent socket, which is dian edges. Distally, the second to fourth exopodal articles less extensive than in the preceding two post−antennular each give rise to a subsequent article and carry one and two limbs; also, the armature on the median edge with setae or setae medio−distally respectively. The fifth exopodal article spines is less well developed. One cluster of setae, including bears 3 setae along its distal margin (indicated by the speci− a large spine−like one, is situated about two−thirds up from men of the earliest stage, UB W 265, see Fig. 9B). the proximal edge of the basipod, and a triplet of setae arises near the medio−distal margin (Fig. 2C). Third cephalic appendage (app3).—As the previous limb Of the endopod, only the proximal podomeres (first one and all following ones, the second post−antennular limb (Fig. and parts of the second one) are known. The proximal podo− 7C) is biramous. It inserts behind the posterior margin of the mere is almost as broad as the basipod and carries a set of hypostome, lateral to the elevation of the sternal area (Fig. setae medially and a row of denticles medio−distally (Fig. 1E). Its basipod is about 30% longer than that of the second 6C1). The outer margin articulates with the second article of cephalic appendage. Its posterior face is shorter than the an− the exopod (Fig. 1F). The second endopodal podomere in− terior. The extensive arthrodial membrane forms the limb serts on the medio−distal end of the proximal podomere and socket. Medially, the proximal endite, a small sclerotised ele− is roughly 30% narrower than the latter. It is too poorly ment, rests within this socket. It is sub−triangular and carries known for further description. a single seta close to its posterior edge. The armature on the The exopod consists of two portions. Together they form median edge is principally bi−serial, but the proximal ele− a paddle and are separated by a fine hinge joint, which ex− ments of the series are prominent spines (Fig. 4B). The distal tends from the boundary of the basipod and the first endo− setae are arranged in two distinct rows each with four setae podal podomere to the lateral margin of the exopod (Fig. along the median edge plus two setae on the posterior face 8D). The proximal exopodal portion is subtriangular. Later− (Fig. 5B2). In addition, there is a series of three small setae ally, it is somewhat extended and dipped into the membra− anterior to the larger spines. nous cuticle between shield margin and body proper, form−

Fig. 6. Details of specimens of later developmental stages of the stem crustacean Oelandocaris oelandica Müller, 1983 from Gum continued. A. UB W 263, ® close−up median view of appendages 4 to 6 exhibiting denticle rows on the distal area of the podomeres (arrows). B. UB W 260; B1, lateral close−up view of specimen exhibiting appendages posterior to the third cephalic appendage; note the exopod shaft being articulating with body wall (arrows); B2, posterior to ventral view of strongly bent specimen; tips of limbs broken off rather proximally. C. UB W 261; C1, close−up lateral view of the exopods of appendages five and six displaying putative attachment sites of delicate setae or sensillae (arrows), and such sensillae at the tergite (arrow heads); C2, dorsal view of trunk segments; arrows point to attachment sites of delicate setae or sensillae; C3, dorso−lateral and close−up view of the posterior tergite displayed in C2; arrows point to attachment sites of delicate setae or sensillae. Abbreviation: an, anus. STEIN ET AL.—CAMBRIAN STEM CRUSTACEAN 473

http://app.pan.pl/acta53/app53−461.pdf 474 ACTA PALAEONTOLOGICA POLONICA 53 (3), 2008

am

am peduncle article 1 + outgrowth 1 peduncle article 2 + outgrowth 2

peduncle article 3 + outgrowth 3

basipod endopod

exopod

proximal endite

Fig. 7. Reconstruction of the appendage morphology of later developmental stages of Oelandocaris oelandica Müller, 1983. A. Antennula. B–F. Second to sixth appendages. Note that the first three appendages differ from each other and from the sub−equal fourth to sixth appendages. ing a monocondylic limb joint (Fig. 2E2). Along almost the Trunk appendages.—The anterior trunk appendages are entire margin, the paddle carries prominent setae intercalat− known to some degree (Figs. 2A, C, E1, 7F), whereas the pos− ing with finer setae (reconstructed in Fig. 7D–E, see also Fig. terior ones are known only from fragments of their proximal 8A). In addition, the lateral edge of the proximal triangular parts. From what is preserved, it seems that the anterior ones portion carries one seta far distally (known by its socket; Fig. are similar to the third and fourth post−antennular limbs, but 8D). The outer rim of the exopod appears more strongly have progressively narrower basipods and decrease in over− sclerotised than the anterior and posterior surfaces (possibly all size towards the posterior. Additionally, the space be− stabilizing the rather soft ramus; arrowed in Fig. 8D). tween the basipods of each appendage pair becomes progres− STEIN ET AL.—CAMBRIAN STEM CRUSTACEAN 475

arthrodial membrane seta

seta

Fig. 8. SEM photographs of additional important details of the stem crustacean Oelandocaris oelandica Müller, 1983. A. UB W 262, exopods of append− ages posterior to the third cephalic appendage, showing insertion of setae; note the weak sclerotisation of the exopod surface as demonstrated by the de− formed exopod of the fifth appendage. B. UB W 261; B1, tail piece with marginal spines on last tergite and caudal end (arrows); B2, anterior wings of hypostome (arrow); B3, insertion sites of sensillae on tergite borders and exopods (arrows). C. UB 649, anal opening with displaced membrane on ventral side of caudal end. D. UB W 260, basipods and exopods of third to sixth limbs, showing the more strongly sclerotised lateral margin of the exopod (arrows) with seta distally, and the membranous area at the articulation of the exopod of the third limb with the basipod. Abbreviations: alt, antennula; an, anus; app, appendages; bas, basipod; ex, exopod; hyp, hypostome; me, median eyes; tel, telson; ts, tergites.

http://app.pan.pl/acta53/app53−461.pdf 476 ACTA PALAEONTOLOGICA POLONICA 53 (3), 2008 sively narrower towards the posterior, and the setal armature Denticles.—Tiny cuticular spines, called denticles, occur in becomes less developed (Figs. 2E1, 8D). rows along the mediodistal edge of the first endopodal podo− meres of the fourth and fifth appendage and all post−cephalic limbs (Fig. 6A). Denticles are otherwise known so far only Special morphological details from labrophoran crustaceans (see, e.g., Müller and Walos− Setation.—Preservation of setae and spines in the available sek 1988: pl. 4: 3), but the recent report from the stem material is incomplete. Prominent and possibly large setae or chelicerate Leanchoilia illecebrosa (Hou, 1987) by Liu et al. spines have a higher preservational potential than delicate (2007) suggests that they possibly represent a rather ancient ones, but often are broken off distally. In most cases, only the feature. sockets of the more delicate setae are preserved. Therefore, reconstruction of the setation of Oelandocaris oelandica re− Ontogeny lies largely on size, form and distribution of these sockets and Early developmental stage.—One specimen, UB W 265, experience from other “Orsten” taxa. considered to belong to Oelandocaris oelandica, is consider− There are, at least, three kinds of sockets to be distin− ably smaller (440 μm total length, reconstruction in ventral as− guished pointing to the presence of different setae or spines pect in Fig. 10A) than the other specimens (660 μm to more arising from these. One is an extension of the general cuti− than 990 μm total length). Its head shield (Fig. 9A, B) incorpo− cle. This type is the predominant type found, e.g., along the rates five appendiferous segments as in the larger specimens, median edges of the basipods, at least some possibly having has similarly little extended margins, but has a more oval masticatory function (Fig. 5B , C). A second type is a 2 shape in dorsal view, with a truncated posterior midline. The socket at the edge of the exopod flaps of the fourth and suc− hypostome (Fig. 9B, C, E; mouth in F) differs in shape from in cessive limbs. These suggest the insertion of long and the larger specimens in being more oval in ventral view and prominent setae, often used for locomotion (Fig. 5A ). An− 1 posteriorly pointed. Another difference is in the antennula, other, most likely more delicate type of setae occurs on the which has presumably only two outgrowths arising from the antennular outgrowths (Fig. 4B), the median edges of the proximal part (Fig. 9G). Assignment to Oelandocaris oelan− basipods, and along the inner edges of the exopods oppos− dica is supported by the paired structure overlapping the ante− ing the endopods (Fig. 5B ). Those on the inner edges of the 1 rior wings of the hypostome (Fig. 9D, E as compared with Fig. basipods are preserved only rarely and never in full length. 8B ), multi−annulated exopods of second and third cephalic The function of this type of setae is unclear; a function as 2 appendages with the terminal portion of the third cephalic ap− part of the feeding apparatus is possible, though not obvious pendage carrying three setae (Fig. 9G) and paddle−shaped for those setae on the antennular outgrowths. A third type of exopods of the posterior limbs having robust marginal setae, sockets found on the tergites and outer edges of the proxi− although fewer in number (Fig. 9H). A further indication of its mal exopod podomeres is a much smaller one, having a cir− inclusion into this species is the presence of a lateral step−like cular depression in the centre (Fig. 6C; for details see next ridge on either side of the conical, slightly depressed tail end section). representing the insertion points of, most likely, postero−later− Possible sensilla.—The tergites are covered with holes of ally pointing spines (arrows in Fig. 9B, C). The divergent mor− phology of the hypostome is bridged by the next larger speci− about 5 μm in diameter (Fig. 6C1,C2). In some cases a re− men UB W 262 (more than 670 μm total length), which pos− cessed central socket is preserved (Fig. 6C3), suggestive of a sensillum arising there. Such holes occur in rows parallel to sesses a hypostome of intermediate shape. the anterior and posterior margins of each tergite. Additional Apart from its size, the specimen differs from later devel− sockets occur in a row along the lateral margins of the ter− opmental stages in the: – morphology of the head shield; gites (Fig. 6C1). Some sockets in this row are larger, particu− larly the most posterior which probably give rise to a more – antennula that has only two outgrowths; prominent sensillum or seta. Several similarly small pores, – basipod morphology; filled by a membrane except for a tiny hole in the centre – shape of the trunk; – number of trunk segments. (Figs. 6C1,8B3), are located on the anterior surface of the tri− angular exopod portion and the proximal part of the pad− Morphology of the early stage in detail.—The specimen dle−shaped distal exopod portion close to the lateral margin. has a bowl−shaped head shield incorporating five appendi− These may have been the insertions of fine sensilla, such as ferous segments. Anteriorly, the head shield margin appears are known from “Orsten” and extant eucrustaceans. to be amply rounded (Fig. 9A–C), but it is deformed in the

Fig. 9. Earliest known developmental stage of the stem crustacean Oelandocaris oelandica Müller, 1983, specimen UB W 265 from Gum, Västergötland, ® Sweden. A. Dorsal view. B. Lateral view, arrow marks step−like ridge on tail, arrow head marking small seta. C. Ventral view, arrows mark step−like ridges on tail. D. Close−up ventral view of the paired structures interpreted as median eyes. E. Close−up ventral view of the forehead, arrow marks posterior wing of hypostome. F. View into the mouth opening. Note the possible sensory structures surrounding it (arrows). G. Latero−ventral view of posterior head region. H. Close−up posterior view of posterior limbs. Abbreviations: app, appendages; me, median eyes; ex, exopod; hyp, hypostome. STEIN ET AL.—CAMBRIAN STEM CRUSTACEAN 477

head shield tergite 1

head shield

tergite 1

antennula

median eyes

mouth

http://app.pan.pl/acta53/app53−461.pdf 478 ACTA PALAEONTOLOGICA POLONICA 53 (3), 2008

median eyes hypostome antennula

app2

app3

app4

Fig. 10. 3D model of the morphology of the early developmental stage and the ontogenetic sequence of Oelandocaris oelandica, with four developmental stages known. A. Early developmental stage in ventral view, for comparison, but not to scale the ventral view of the oldest developmental stage known is shown. B. Reconstruction of stages in lateral view. Early developmental stage (B1), first next older set (B2), second next older set (B3), and third, apparently, oldest stage (B4). Abbreviations: app, appendages. STEIN ET AL.—CAMBRIAN STEM CRUSTACEAN 479 single specimen at hand. A rostrum−like protrusion of the ments with one stout seta medio−distally (Fig. 9E). The shield, as in the later stages, is lacking. The trunk is a single, exopods of the following two appendages are paddle−shaped dorso−ventrally slightly depressed cone−shaped portion, but as in the larger specimens, but differ from those of the older must include at least one segment, as is indicated by the sixth ones in that the paddles are made up of a single portion only. pair of appendages inserting ventrally (Fig. 9B). The future Furthermore, the outer joint is formed by the basipod and it tergite of this trunk segment, laying partly underneath the seems that the first endopod podomere of the fourth limb is posterior border of the head shield, is also indicated by the not articulating with the exopod. The exopod of the third presence of a small spine at its weakly demarcated postero− post−antennular limbs carries six setae along its margin, two lateral corners (arrowhead in Fig. 9B), similar to the situation inserting disto−medially, one terminally (the most prominent in the head shield and tergites of the larger specimens. The one), and three disto−laterally. A smaller associate seta is trunk is approximately half as long as the head shield. Poste− present antero−laterally to the terminal seta. The exopod of riorly, it bends slightly dorsally. Laterally in this part of the the fourth post−antennular limb carries five setae along its trunk there is a step−like ridge on either side at about three− margin, one inserting disto−medially, one terminally (the quarters of the length of the trunk. These ridges form the in− most prominent one) and three disto−laterally. The single pair sertion points of fine, most likely postero−laterally projecting of trunk limbs is almost completely hidden under the exo− setae or spines (Fig. 9B, C). pods of the fifth appendage. It appears to be uniramous, car− The main body of the hypostome is ovoid, with a wider rying three setae distally, the anterior one being smaller than front and a tapering rear. The pointed posterior midline bore a the other two. This seta may be the precursor of the endopod. spine, which is broken off in the specimen at hand (Fig. 9E). This spine is flanked on either side by a spine arising from the Later developmental stages, i.e,. rostrum−bearing instars. posterior margin of the hypostome. The mouth opening is visi− —The six larger specimens (670–990 μm) all have head ble below the posterior of the hypostome (Fig. 9F). The ante− shields that incorporate five appendiferous segments, a ros− rior wings of the hypostome are overlain by the paired bulbous trum−like anterior projection of the shield, and three to five structure (Fig. 9D, E) also known from the larger specimens trunk segments free from the tailpiece. Possibly three (ros− and tentatively interpreted there as median eyes. Due to some trum−bearing) stages can be identified: degree of deformation, the insertion of the right antennula is – smallest stage: total length slightly longer than 670 μm, contorted and the posterior wing of the hypostome pushed an− three tergites (UB W 262; the total length can not be mea− teriorly (arrowed in Fig. 9E). The location of the second ce− sured, as the anterior part of the head shield and posterior phalic appendage indicates that the original position of the pos− portion of the tail are missing); terior wings was farther back, about two−thirds the length of the – median stage: total length estimated to be between 680 and hypostome. Accordingly, only the pointed posterior end of the 740 μm, four tergites (UB 648, holotype, UB W 264; total hypostome slightly overhung part of the postoral sternal area. length of the latter can not be measured, as the anterior The left antennula is largely complete. Its shows an ample portion of the head is missing); arthrodial membrane, which is rather squeezed together, and – largest stage: total length almost 1mm, five tergites (UB W several tubular articles. The first article bears a short spine or 260, 261, 263). seta antero−distally on a slight swelling of the margin (Fig. The difference in size and number of tergites between the 9E), the second gives rise to a tubular, but preservationally smallest rostrum−bearing instar, having three pairs of fully de− flattened outgrowth latero−distally, while the third article veloped limbs posterior to the third cephalic appendage, and gives rise to two more portions, the proximal one carrying a the early larva UB W 265 indicates a gap in preserved seta antero−distally and the distal portion being a rounded ontogenetic stages (Fig. 10B). The most apparent difference cone, possibly continuing into a terminal seta (Fig. 9G). This between the smallest of the rostrum−bearing instars and the morphology resembles that of the antennula of older stages, two larger rostrum−bearing instars is that the posterior end of suggesting that the proximal outgrowth is still missing— the hypostome is pointed (Fig. 2D), as in the early develop− possibly the seta on the first article, while the second is pres− mental stage (Fig. 9E). The head shield of the holotype has a ent, and the third outgrowth consisting of the distal portions weak boundary between the fourth and fifth appendiferous of the antennula (Fig. 9E). segments, which cannot be seen in UB W 264, the other speci− In all post−antennular limbs, setation of the basipod of the men of this instar. In the smaller UB W 262, the head shield is small specimen is less developed than in the larger speci− broken dorsally and the remainder is bent in several places. mens. A “proximal endite” could not be observed. The exo− Therefore, it cannot be assessed whether this demarcation is podal articles of the second and third cephalic appendages in− present also in the smallest rostrum−bearing instar. As can be crease in length from proximal to distal and are, in all, more judged from the limited material, morphological differences cylindrical (longer than wide) than those of the same limbs in within the largest two instars reflect a further increase in size larger stages (Fig. 9B, C, G). Only two exopodal portions and and a larger number of tergites (5), the latter feature being the two endopodal podomeres are preserved of the third cephalic most convincing feature for stage discrimination. All three in− appendage. The exopodal portions lack setae on their inner stars can be clearly ordered or distinguished by the progressive margins (Fig. 9B, C, G). The endopods are cylindrical ele− increase in the number of free trunk segments.

http://app.pan.pl/acta53/app53−461.pdf 480 ACTA PALAEONTOLOGICA POLONICA 53 (3), 2008

“Orsten” material. The species is interpreted as a eucrustacean Discussion since it shares various features with cephalocarids, branchio− pods and “Orsten” crustaceans R. kinnekullensis, Dala pei− Morphological features lertae and W. quinquespinosa (Zhang et al. 2007). All are con− sidered as members of the Eucrustacea (cf. Walossek 2003b). Eyes.—Early and later stages of Oelandocaris oelandica,or Y. dianensis has a pair of immense lobes extending from the more precisely the largest instar at hand, possess a pair of bul− forehead in early stages, similar to those of R. kinnekullensis, bous structures at the front of their hypostomes. Such struc− and a smaller median lobe slightly more posteriorly located tures are unknown from any fossil other than in “Orsten” type than the pair. of preservation or extant euarthropods and their larvae. Com− Two species cannot be compared in detail at present: the parisons with fossil taxa are, therefore, limited, more or less, to fragmentary head preservation does not allow assessment of “Orsten” taxa. Among these, several taxa bear paired frontal, eyes in Dala peilertae Müller, 1983. The material known of lobe−like structures, at least in early stages of their develop− Cambrocaris baltica Walossek and Szaniawski, 1991 is too ment. Within Eucrustacea these are Bredocaris admirabilis coarsely preserved (Walossek and Szaniawski 1991). Unfor− Müller, 1983, Rehbachiella kinnekullensis Müller, 1983, and tunately, the specimen is now destroyed and therefore cannot Walossekia quinquespinosa Müller, 1983. The lobes of B. be re−scanned. admirabilis are fairly large and separated by a bar with a me− In all, only a few of the mentioned eye−like structures dian pit or pimple−like structure (Müller and Walossek 1988: could be demonstrated to be compound eyes, e.g., on the ba− pl. 8:2). R. kinnekullensis has a pair of large lobes initially sis of characteristic facet patterns of the original ommatidia which can be followed during ontogeny (Walossek 1993: pls. (although this cannot be visualised in an SEM, and even 21: 1, 28: 4). Similar to B. admirabilis, there is a median struc− present ommatidia must not necessarily display a facet pat− ture, but more bulging than in the latter. W. quinquespinosa tern on the surface), an elaborate eye−stalk that is unknown has a pair of rather pointed and closely spaced lobes at the for median eyes, or less so by the topology. With respect to front of the hypostome/labrum complex (Müller 1983: “fo” in O. oelandica, it seems appropriate, therefore, to follow his fig. 6) which are similar to those of O. oelandica. Müller (1983) and Stein et al. (2005) in interpreting the Phosphatocopina, the sister taxon of Eucrustacea accord− paired structure at the anterior end of its hypostome as a the ing to Siveter et al. (2003) and Maas et al. (2003), is charac− median (naupliar) eye rather than as vestiges or rudiments of terised by paired lobes with probably an associated third part the lateral compound (= facetted) eyes. Thus, compound posterior to the lobes on the anterior part of the hypostome, eyes are evident only for the “Orsten” taxa Henningsmoeni− interpreted as median eyes because of their position (Maas et caris scutula and the two one−eyed species Goticaris longi− al. 2003: fig. 3A–D). Within the stem−lineage derivatives of spinosa and Cambropachycope clarksoni. Occurrence of Labrophora recovered from the “Orsten” (Waloszek 2003a), compound eyes, however, is a plesiomorphic character since a pair of lobes, possibly median eyes, is located between the they are part already of the ground pattern of Arthropoda unpaired compound eye and forehead in Goticaris longi− sensu stricto (Waloszek et al. 2005). spinosa Walossek and Müller, 1990 (Walossek and Müller 1990: fig. 3A, B); Cambropachycope clarksoni Walossek Head segments.—A feature that requires further investiga− and Müller, 1990 only possesses the compound eye (Walos− tion in other taxa derived from the stem lineage of the sek and Müller 1990). A pair of lobes of Henningsmoeni− Labrophora is the number of appendiferous segments in the caris scutula (Walossek and Müller, 1990) rests on stalks and head. The holotype (UB 648) of Oelandocaris oelandica has been interpreted as the compound eyes (Walossek and shows a weak boundary between the fourth and fifth ap− Müller 1990). Other “Orsten” taxa have been described as pendiferous head segments (see above). The reasonably well eyeless, such as the eucrustacean skaracarids (Müller and preserved head shields of other specimens (UB W 264 of the Walossek 1985b) and the euarthropod Agnostus pisiformis same stage as the holotype and UB W 263 representing the (Wahlenberg, 1818) (Müller and Walossek 1987), but the latest stage at hand) do not show this split, suggesting that the two have indeed paired lobes in the frontal area: in Skara line on the holotype may be interpreted as a preservational anulata these occur antero−ventrally on the hypostome− feature. From a phylogenetic perspective it is clear, though, labrum complex (Müller and Walossek 1985b: pl. 4: 2, 3), that the fifth appendiferous segment was added independ− and A. pisiformis possesses small humps on the pliable cutic− ently in the different euarthropod lineages. One instance oc− ular area in front of the hypostome (Müller and Walossek curred in the stem lineage at least of the Labrophora, another 1987: pl. 13). in the chelicerate stem−lineage. Another euarthropod species recently discovered in China The addition of original trunk segments to a head tagma is Yicaris dianensis Zhang, Siveter, Walossek, and Maas, can be recognised during the ontogeny of Eucrustacea as dem− 2007 from the Lower Cambrian of Xiaotan section, Yong− onstrated by their larvae attaining more segments progres− shan, Yunnan Province (Zhang et al. 2007). Y. dianensis is sively. Ontogeny within Eucrustacea starts with a nauplius known from material preserved in “Orsten” type of preserva− larva with three appendiferous segments. This feature is re− tion and, therefore, is easily comparable with the other garded as an autapomorphy of the taxon (Maas et al. 2003; STEIN ET AL.—CAMBRIAN STEM CRUSTACEAN 481

Waloszek 2003a). Among Recent eucrustaceans, the smallest oelandica. Further detailed investigation on other stem−lin− metanauplius has four appendiferous segments incorporated eage derivatives of the Labrophora is needed to clarify this in the head, and this status is retained for a period of stages un− matter. til the fifth segment is included seamlessly (cf. Müller and Furthermore, it is established that, not before the ground Walossek 1988 for Bredocaris admirabilis; Walossek 1993 pattern of the Labrophora, the proximal endites of the an− for Rehbachiella kinnekullensis;JensHøeg, personal commu− tenna and the mandible autapomorphically became modified nication 2003 for cirripede cyprid larvae). into coxae, forming a ring−shaped structure underneath and Yet, ontogenetic and evolutionary changes have to be re− articulating against the basipod. Medially, this ring−shaped garded separately. So far, only Martinssonia elongata Mül− structure is drawn out into an endite with marginal spines ler and Walossek, 1986 as a representative of the labro− (gnathobase) on the second limb or antenna, which is slightly phoran stem lineage has been investigated in detail (Müller turned against the labrum (Maas et al. 2003; Siveter et al. and Walossek 1986). In this species all instars show a demar− 2003). In Labrophora, all postmandibular limbs bear a proxi− cation of the fifth segment dorsally, while the lateral margin mal endite, as exemplified by the Phosphatocopina (Maas et is continuous. Thus, it remains unclear whether this is a spe− al. 2003) and many extant Eucrustacea (Calman 1909 who cialisation of the fifth head segment in being a demarcation called the proximal endite “arthrite” in his work on branchio− as a joint within the head to lift the anterior part (such kinetics pods; see also Walossek 1993, 1999 and Walossek and occur especially in stem arthropods; cf. Waloszek et al. Müller 1998a, b for an overview of the fate of the proximal 2005) or a retained plesiomorphy with a “not yet” completely endite within Eucrustacea). included status of the segment. Even the largest specimens of Recently, Siveter et al. (2007) described a new arthropod, Cambropachycope clarksoni and Goticaris longispinosa Tanazios dokeron Siveter, Sutton, Briggs, and Siveter, 2007, have no more than four appendiferous segments. Re−study of from the famous Silurian Herefordshire Lagerstätte, UK, and Henningsmoenicaris scutula is under way and it seems clear interpreted the fossil as a putative stem crustacean. However, that the largest specimens have a head showing five appen− the rather coarse preservation of the material with respect to diferous segments as stated by Walossek and Müller (1990). the 3D reconstruction does not allow for the detection of any The presence of a head with five appendiferous segments of the critical features of Crustacea, particularly the “proxi− may therefore be a parallel development of O. oelandica mal endite” or its derivative, the “coxa”. At least, the inter− comparable with the clearly convergent evolution of six or pretations made by Siveter et al. (2007) are not in conflict more−segmented cephalothoracic shields among eucrusta− with our interpretation of O. oelandica as an early offshoot of ceans, but a definitive judgement of this issue has to await the crustacean lineage. further details on the other derivatives of the labrophoran It is of significance that Crustacea sensu lato lack coxae stem lineage. on what are called “antennae” and “mandibles” in labro− phoran crustaceans, terms that we would rather restrictedly “Proximal endite”.—The “proximal endite” as a movable, use for this evolutionary level. Consequently, there was no setose element in the arthrodial membrane medially under− coxal gnathobase for mastication on the third cephalic ap− neath the basipod of the postantennular limbs has been re− pendage early in the crustacean lineage, which developed garded as an autapomorphy for Crustacea sensu lato since only subsequently in the stem species of the Labrophora. It 1990 (Walossek and Müller 1990; Walossek 1999; Maas is also clear that O. oelandica (and the rest of the labro− and Waloszek 2003; Waloszek 2003a). After the re−study of phoran stem−lineage taxa) is not simply a euarthropod, but material from Oelandocaris oelandica by Stein et al. (2005) is associated with the labrophoran stem lineage, in other it became clear that the evolutionary origin of the proximal words is a crustacean sensu lato. Nonetheless, plesiomor− endite is more complicated than hitherto assumed. In fact O. phies such as the hypostome with the mouth at its rear, lack oelandica has only one proximal endite on its third cephalic of a labrum, specialized second and third limbs without appendage (the mandible of labrophoran crustaceans); all coxal portions and serially uniform limbs posterior to the other limbs, the second appendage and more posterior ones, third cephalic appendage clearly determine the basal phylo− lack a proximal endite and there is no evidence that it has genetic position of O. oelandica on the early stem lineage of been lost or modified into another structure. New investiga− Labrophora. tions of known “Orsten” taxa (JTH, AM, and DW unpub− lished data) indicate that the proximal endite may develop Possible life habits of the oldest known instar ontogenetically on a single limb first, and more proximal of Oelandocaris oelandica endites occur progressively on additional limbs during later ontogeny; but it may even not appear on all postantennular The most probable life habit of Oelandocaris oelandica,re− limbs throughout the known ontogeny of “Orsten” fossils ferring to the latest developmental stage known so far, of assigned to stem−lineage derivatives of Labrophora. Phylo− course, seems to have been that of an active swimmer. That genetically, the appearance of the proximal endite seems to assumption is supported mainly by its limb morphology. occur not on all but being restricted to a single or at least a The large paddle−shaped exopods of the limbs posterior to few post−antennular limbs only, as demonstrated by O. the third cephalic appendage seem to be organs well suited

http://app.pan.pl/acta53/app53−461.pdf 482 ACTA PALAEONTOLOGICA POLONICA 53 (3), 2008

stroke of the exopods. The extensive arthrodial membrane forming limb sockets increased agility of the limb as a whole. It seems that the limb as a whole was swung back− wards during the swimming stroke. The subdivision of the exopod probably enabled reduction of drag during the re− covery stroke by bending the distal portions backwards. The basic posture of the large antennula clearly was di− rected postero−ventrally. Given posture and articulation of the antennula, a sweeping movement seems most plausible, during which it was swung backwards from a ventrally di− rected starting position during stroke. In that movement, the antennula would have raked food particles into the ventral food grove. During the recovery stroke it would have moved back into the starting position. This would pose problems for bottom living habits, where the sediment sur− face would hinder such a movement. Even in a potential posteriorly directed resting position, the antennula would have interfered with the endopods of the post−antennular limbs. It seems therefore most reasonable to assume that O. oelandica, at least up to the latest stage known so far, ac− tively swam in the water column and, if closer to the bot− tom, possibly upside down, as many small crustaceans do, namely brine shrimps, fairy shrimps or tadpole shrimps. We assume that all limbs acted in concert in a meta− chronal movement; a putative sequence of movement is pre− sented in Fig. 11. During the stroke, the postantennular limbs are moved outward and backwards, opening the ventral food groove. Simultaneously, the antennula was swung back, ini− tially spreading the long outgrowths, to sweep a large area. Towards the end of the stroke sequence, the outgrowths are drawn together to allow sweeping down into the ventral groove. During the recovery stroke, the antennula swings an− teriorly, in the posteroventral starting position. The post− antennular limbs simultaneously swing back anteriorly and inwardly. The setation of the basipods then functions in transporting food particles anteriorly towards the mouth opening. The inward facing setae on the exopods of the sec− ond and third postantennular limbs may have functioned in retaining escaping particles in the general cephalic area, to be brought in the food groove with the next stroke.

Fig. 11. Reconstruction of sequence of appendage−movement of the stem Conclusions crustacean Oelandocaris oelandica Müller, 1983, captured in succeeding steps from above towards below. Please watch movie at data repository of Refining the description of Oelandocaris oelandica much of Acta Palaeontologica Polonica (http://app.pan.pl/SOM/app53−Stein_etal_ the data presented by Stein et al. (2005) could be confirmed. SOM.pdf). In addition we can identify the presence of four different in− stars even in the limited material. This significant new obser− for locomotion. Such exopods are possibly retained from vation indicated by size increase between the stages and a the euarthropod ground pattern (cf. Chen et al. 2004; stepwise addition of trunk tergites (reflecting somite addi− Waloszek et al. 2005). The connection between the proxi− tion) from one in the earliest larva−like stage to three, four mal endopodal podomere and the exopod, possibly even an− and five in the rostrum−bearing stages allowed also for an ac− other plesiomorphy retained from the euarthropod ground count of aspects of ontogeny. During growth few other de− pattern (see Liu et al. 2007 for the stem chelicerate tails are modified significantly, enabling a combination of all Leanchoilia illecebrosa), both reduced the flexibility of the information into a reconstruction of the general morphology endopods and enhanced the ability for a more forceful of the species (Fig. 3). Importantly, no more proximal endites STEIN ET AL.—CAMBRIAN STEM CRUSTACEAN 483 than the one on the third cephalic appendage are added up to the latest stage identified so far. Acknowledgements O. The new data revealed by the new specimens of Most of the images were made at the Central Facility for Electron Mi− oelandica, confirm the view of Stein et al. (2005) that this croscopy of the University of Ulm. Jens Høeg (University of Copenha− “Orsten” species cannot have branched off from the evolu− gen, Copenhagen, Denmark) is thanked for comments on cirripede lar− tionary lineage of the Crustacea after, e.g., Cambropachy− vae. John S. Peel (Uppsala University, Uppsala, Sweden), and John E. cope clarksoni, Goticaris longispinosa, Henningsmoenicaris Repetski (U.S. Geological Survey, Reston, VA, USA), kindly reviewed scutula, and Martinssonia elongata. There is growing evi− the manuscript and improved the language. We are also grateful to dence that in early developmental stages some of the other Geoff Boxshall (Natural History Museum, London, UK) for valuable comments. The plasticine model was built with the skilful help of two stem crustaceans lack the “proximal endite” on appendages pupils from the Humboldt Gymnasium in Ulm, Marie Holdick and other than the third one. Accordingly, the appearance and Florian Rassmann, during a “practical week” at our section. This is a specific fate of the proximal endite may prove to be eventu− contribution from the Center of “Orsten” Research & Exploration ally even a significant tool to resolve the interrelationships C.O.R.E. [www.core−orsten−research.de]. between the stem crustaceans, leading to a reconstruction of the beginning of the crustacean evolutionary lineage. Such a step awaits the detailed description of all available taxa but References the data available already now help to firmly fix the position of O. oelandica on the stem lineage toward the labrophoran Babcock, L.E., Peng S.C., Geyer, G., and Shergold, J.H. 2005. Changing Crustacea (cf. Siveter et al. 2003), as suggested already by perspectives on Cambrian chronostratigraphy and progress toward sub− Stein et al. (2005). This interpretation is supported by the: division of the Cambrian System. Geosciences Journal 9: 101–106. – presence of at least one proximal endite; Budd, G.E. 2002. A palaeontological solution to the arthropod head prob− lem. Nature 417: 271–275. – exopods of the second and third cephalic appendages be− Calman, W.T. 1909. Appendiculata. Vol. 3. Crustacea. In: R. Lankester ing multi−annulated and having medially oriented setae; (ed.), A Treatise on Zoology, 1–346. Adam & Charles Black, London. – presence of a bi−partite locomotive apparatus consisting of Chen, J., Waloszek, D., and Maas, A. 2004. A new “great appendage” ar− an anterior set, the antennulae acting in conjunction with thropod from the Lower Cambrian of China and the phylogeny of the subsequent two pairs of head appendages, and a poste− Chelicerata. Lethaia 37: 3–20. Chen, J., Waloszek, D., Maas, A., Braun, A., Huang, D., Wang, X., and rior set of more or less homonomous limbs working to− Stein, M. 2007. Early Cambrian Yangtze Plate Maotianshan Shale gether in a metachronal beat. macrofauna biodiversity and the evolution of predation. Palaeogeo− Again, the long and prominent antennula with its three graphy, Palaeoclimatology, Palaeoecology 254: 250–272. rod−like outgrowths of O. oelandica was clearly not a sen− Glenner, H., Thomsen, P.F., Hebsgaard, M.B., Sørensen, M.V., and Wil− sorial device, but was, most likely, used for food gathering, lerslev, E. 2006. The Origin of Insects. Science 314: 1883–1884. Liu, Y., Hou, X., and Bergström, J. 2007. Chengjiang arthropod Leanchoilia as suggested already by Stein et al. (2005). This supports in− illecebrosa (Hou, 1987) reconsidered. GFF 129: 263–272. terpretations of the primary function of the antennula al− Maas, A. and Waloszek, D. 2005. Phosphatocopina—ostracode like sister ready in the crustacean ground pattern as a food−gathering group of Eucrustacea. Hydrobiologia 538: 139–152. device (Waloszek et al. 2005, 2007). Assumptions concern− Maas, A., Waloszek, D., and Müller, K.J. 2003. Morphology, ontogeny and ing the relationships of insects and myriapods, both taxa phylogeny of the Phosphatocopina (Crustacea) from the Upper Cam− bearing feeler−like antennulae, with the Crustacea sensu brian “Orsten” of Sweden. Fossils and Strata 49: 1–238. Maas, A., Braun, A., Dong, X., Donoghue, P., Müller, K.J., Olempska, E., lato, the Labrophora or any of the eucrustacean in−groups Repetski, J.E., Siveter, D.J., Stein, M., and Waloszek, D. 2006. The (e.g., Glenner et al. 2006) must take this into consideration. “Orsten”—more than a Cambrian Konservat−Lagerstätte yielding ex− If advocating the assumption that insects are crustaceans, ceptional preservation. Palaeoworld 15: 266–282. one has to explain at which node (in which stem species) Müller, K.J. 1983. Crustacea with preserved soft parts from the Upper Cam− particular similarities occurred (developed) to be trans− brian of Sweden. Lethaia 16: 93–109. Müller, K.J. and Walossek, D. 1985a. A remarkable arthropod fauna from ferred into synapomorphies of particularly insects and a the Upper Cambrian “Orsten” of Sweden. Transactions of the Royal So− specific crustacean in−group. It also needs to be explained ciety of Edinburgh: Earth Sciences 76: 161–172. when and how specific modifications (autapomorphies) oc− Müller, K.J. and Walossek, D. 1985b. Skaracarida, a new order of Crustacea curred in the specific lineage toward the insects and the ac− from the Upper Cambrian of Västergötland, Sweden. Fossils and Strata cording eucrustacean in−group. No fossils are available as 17: 1–65. yet to bridge gaps for any of these lineage sections. Again, Müller, K.J. and Walossek, D. 1986. Martinssonia elongata gen. et sp. n., a crustacean−like euarthropod from the Upper Cambrian of Sweden. the fate of features that crustaceans retained from earlier Zoologica Scripta 15: 73–92. nodes and features that are acquired by the stem species of Müller, K.J. and Walossek, D. 1987. Morphology, ontogeny, and life−habit Crustacea sensu lato, Labrophora and Eucrustacea have to of Agnostus pisiformis from the Upper Cambrian of Sweden. Fossils be explained for crustacean in−group taxa. That means one and Strata 19: 1–124. needs to give reason for the lack of such features in insects Müller, K.J. and Walossek, D. 1988. External morphology and larval devel− opment of the Upper Cambrian maxillopod Bredocaris admirabilis. if one considers them to be Eucrustacea. Until then, rela− Fossils and Strata 23: 1–70. tionship hypotheses about insects, myriapods and crusta− Peng, S.C., Babcock, L.E., Robison, R.A., Lin, H.L., Rees, M.N., and Saltz− ceans remain, at best, uncertain. man, M.R. 2004. Global Standard Stratotype−section and Point (GSSP)

http://app.pan.pl/acta53/app53−461.pdf 484 ACTA PALAEONTOLOGICA POLONICA 53 (3), 2008

of the Furongian Series and Paibian Stage (Cambrian). Lethaia 37: Cambrian of Sweden and their bearing upon the position of Agnostus. 365–379. Lethaia 23: 409–427. Scholtz, G. and Edgecombe, G.D. 2005. Heads, Hox and the phylogenetic po− Walossek, D. and Müller, K.J. 1998a. Cambrian “Orsten”−type arthropods and sition of trilobites. In: S. Koenemann and R.A. Jenner (eds.), Crustacea the phylogeny of Crustacea. In: R.A. Fortey and R.H. Thomas (eds.), and Arthropod Relationships, Crustacea Issues 16: 139–165. Taylor & Arthropod Relationships. Systematics Association Special Volume Series Francis, Boca Raton. 55: 139–153. Chapman and Hall, London. Scholtz, G. and Edgecombe, G.D. 2006. The evolution of arthropod heads: Walossek, D. and Müller, K.J. 1998b. Early arthropod phylogeny in the light of reconcilling morphological, developmental and palaeontological evi− the Cambrian “Orsten” fossils. In: G.D. Edgecombe (ed.), Arthropod dence. Development Genes and Evolution 216: 395–415. Fossils and Phylogeny, 185–231. Columbia University Press, New York. Siveter, Da.J., Waloszek, D., and Williams, M. 2003. An Early Cambrian Walossek, D. and Szaniawski, H. 1991. Cambrocaris baltica n. gen. n. sp., phosphatocopid crustacean with three−dimensionally preserved soft a possible stem−lineage crustacean from the Upper Cambrian of Poland. parts from Shropshire, England. In: P.D. Lane, Da.J. Siveter, and R.A. Lethaia 24: 363–378. Fortey (eds.), Trilobites and Their Relatives. Special Papers in Palae− Waloszek, D. 2003a. Cambrian “Orsten”−type arthropods and the phylog− ontology 70: 9–30. eny of Crustacea. In: A. Legakis, S. Sfenthourakis, R. Polymeni, and M. Siveter, De.J., Sutton, M., Briggs, D.E.G., and Siveter, Da.J. 2007. A new Thessalou−Legaki (eds.), The New Panorama of Animal Evolution. Pro− probable stem−lineage crustacean with three−dimensionally preserved ceedings of the 18th International Congress of Zoology, 69–87. Pensoft soft−parts from the Herefordshire (Silurian) Lagerstätte, UK. Proceed− Publishers, Sofia. ings of the Royal Society of London 274: 2099–2107. Waloszek, D. 2003b. The “Orsten” Window—A three−dimensionally pre− Stein, M., Waloszek, D., and Maas, A. 2005. Oelandocaris oelandica and its served Upper Cambrian Meiofauna and its Contribution to our Under− significance to resolving the stemlineage of Crustacea. In: S. Koene− standing of the Evolution of Arthropoda. Paleontological Research 7: mann and R. Vonck (eds.), Crustacea and Arthropod Relationships. 71–88. Crustacean Issues 16: 55–71. Terfelt, F., Ahlberg, P., Eriksson, M.E., and Clarkson, E.N.K. 2005. Furongian Waloszek, D., Chen, J., Maas, A., and Wang. X. 2005. Early Cambrian arthro− (upper Cambrian) biostratigraphy and trilobites of the Håslöv−1 drill core, pods—new insights into arthropod head and structural evolution. Arthro− Scania,S.Sweden.Geologiska Föreningens i Stockholm Förhandlingar pod Structure and Development 34: 189–205. 127: 195–203. Waloszek, D., Maas, A., Chen, J., and Stein, M. 2007. Evolution of cephalic Walossek, D. 1993. The Upper Cambrian Rehbachiella kinnekullensis and feeding structures and the phylogeny of Arthropoda. Palaeogeography, the phylogeny of Branchiopoda and Crustacea. Fossils and Strata 32: Palaeoclimatology, Palaeoecology 254: 273–287. 1–202. Wolff, C. and Scholtz, G. 2006. Cell lineage analysis of the mandibular seg− Walossek, D. 1999. On the Cambrian diversity of Crustacea. In:F.R.Schram ment of the amphipod Orchestia cavimana reveals that the crustacean and J.C. von Vaupel Klein (eds.), Crustaceans and the Biodiversity Crisis. paragnaths are sternal outgrowths and not limbs. Frontiers in Zoology Proceedings of the Fourth International Crustacean Congress, Amster− 2006: 3–19. dam, The Netherlands, July 20–24, 1998, Vol. 1: 3–27. Brill Academic Zhang, X., Siveter, Da.J., Waloszek, D., and Maas, A. 2007. An epipodite− Publishers, Leiden. bearing crown−group crustacean from the Lower Cambrian. Nature Walossek, D. and Müller, K.J. 1990. Stem−lineage crustaceans from the Upper 449: 595–598.